The Myth of Air Power in the Persian Gulf War and the Future of Warfare

The Myth of Air Power Daryl G. Press in the Persian Gulf War and the Future of Warfare

The Myth of Air Power in the Persian Gulf War On February 24, 1991, U.S. groundtroops, supportedby British, French, andArab forces, moved 1 north from Saudi Arabia to liberate Kuwait and destroy the Iraqi military. Four days later the job was apparently done: The Kuwaiti ºag once again ºew over Kuwait City, andwhat remainedof the Iraqi military was in full retreat. Astonishingly, only sixty-three Americans were killedin a groundoperation 2 that most analysts expectedwouldcause thousandsof U.S. casualties. How did the coalition’s ground forces destroy the Iraqi army so quickly and with so few coalition casualties? Does the Persian Gulf War heralda future of U.S. mili - tary dominance and low-cost U.S. military operations? What are the lessons from the ªghting for U.S. foreign policy? The conventional wisdom among historians, military analysts, and foreign policy decisionmakers is that air power neutralized the Iraqi military before 3 the groundwar began. This interpretation of the Gulf War has important im-

Daryl G. Press is Assistant Professor in the Government Department and a Research Fellow at the Rockefeller Center at Dartmouth College. He is also an Associate of the John M. Olin Institute for Strategic Studies at the Weatherhead Center for International Affairs at Harvard University.

The author would like to thank Stephen Biddle, Stephen Brooks, Daniel Byman, Eliot Cohen, Eu- gene Gholz, David Kang, Thomas Keaney, Jennifer Lind, Thomas Mahnken, Michael Mastanduno, Karl Mueller, Kenneth Pollack, Barry Posen, Joshua Spero, Allan Stam, andWilliam Wohlforth for helpful comments on earlier drafts of this article. Previous versions of this article were presented at the Dartmouth Seminar on International Politics, the John M. Olin Institute for Strategic Studies, the StanfordUniversity Center for International Security andCooperation, the Columbia Univer - sity Institute of War andPeace Studies,andthe New York Military Affairs Symposium.

1. Thirty-six countries joinedthe coalition against Iraq, but the groundoffensive can be fairly char - acterizedas a U.S. operation with British support. The coalition’s groundforce comprisednine and two-thirds U.S. divisions, two-thirds of a British division (two brigades), and two-thirds of a French division (two brigades), along with several less-capable Arab divisions. Only the U.S. and British forces were given important roles in the offensive. 2. The number of U.S. soldiersandmarines killedduringthe groundwar is from Stephen T. Hosmer, Psychological Effects of U.S. Air Operations in Four Wars, 1941–1991 (Santa Monica, Calif.: RAND, 1996), p. 155. This number excludesthe twenty-eight Americans killedby a Scudmissile that hit a barracks in Dhahran, Saudi Arabia, during the ground war. For an account of prewar estimates of U.S. casualties, see Benjamin S. Lambeth, The Transformation of American Air Power (Ithaca, N.Y.: Cornell University Press, 2000), p. 2. 3. For the most prominent analyses that support this view, see Eliot A. Cohen, Director, Gulf War Air Power Survey (Washington, D.C.: U.S. Government Printing Ofªce [GPO], 1993), 5 vols. (herein- after citedas GWAPS), plus Summary Report. See speciªcally GWAPS Summary Report, pp. 116–117;

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 6

plications for U.S. foreign policy andfor U.S. military procurement strategies. For U.S. foreign policy, the Gulf War seems to show—andthe 1999 Kosovo conºict appears to conªrm—that air power is now so lethal, andAmerican air power so dominant, that the United States can win nearly cost-free military victories against its foes. For U.S. military procurement debates, the lessons are equally clear: The UnitedStates shouldsigniªcantly alter its military procure - ment priorities to favor air power over groundforces; in the future, decisive 4 battles will be won from the air. U.S. conªdence in air power is soaring, but the conventional wisdom about its decisiveness in the Gulf War has never been rigorously tested. It is undeniable that for six weeks—during the period now known as the air campaign— coalition aircraft droppedtons of bombs andmissiles on Iraqi targets. It is also undeniable that Iraq’s ground forces were totally ineffective against the coalition’s groundforces. But those facts donot prove that the bombing caused Iraqi ineffectiveness. The questions that must be answeredare: Why were the Iraqi groundforces so incapable duringthe groundwar? Were they neutral-

GWAPS, Vol. 2, Pt. 2, p. 107; James A. Winnefeld, Preston Niblack, and Dana J. Johnson, A League of Airmen: U.S. Air Power in the Gulf War (Santa Monica, Calif.: RAND, 1994), pp. 5, 285, 287; Thomas A. Keaney, “The Linkage of Air andGroundPower in the Future of Conºict,” International Security, Vol. 22, No. 2 (Fall 1997), p. 147; Michael R. Gordon and Bernard E. Trainor, The Generals’ War: The Inside Story of the Conºict in the Gulf (Boston: Little, Brown, 1995), p. 474; Rick Atkinson, Crusade: The Untold Story of the Persian Gulf War (Boston: Houghton Mifºin, 1993), p. 495; Hosmer, Psycho- logical Effects of U.S. Air Operations; Robert A. Pape, “The Limits of Precision-Guided Air Power,” Security Studies, Vol. 7, No. 2 (Winter 1997/1998), pp. 93–114; Robert A. Pape, “The Air Force Strikes Back: A Reply to Barry Watts andJohn Warden,” Security Studies, Vol. 7, No. 2 (Winter 1997/1998), pp. 191–214; Aharon Levran, Israeli Strategy after Desert Storm: Lessons of the Second Gulf War (London: Frank Cass, 1997), p. 29. The only academic studies that place substantial emphasis on the importance of the ground campaign are Stephen Biddle, “Victory Misunderstood: What the Gulf War Tells Us about the Future of Conºict,” International Security, Vol. 21, No. 2 (Fall 1996), pp. 139–179; Stephen Biddle, “The Gulf War Debate Redux,” International Security, Vol. 22, No. 2 (Fall 1997), pp. 163–174; and my response to Biddle’s article in Daryl G. Press, “Lessons from GroundCombat in the Gulf: The Impact of Training andTechnology,” ibid.,pp. 137–146. Several U.S. Army-sponsoredaccounts place emphasis on the groundcampaign. See, for example, Robert H. Scales, Certain Victory: The U.S. Army in the Gulf War (Washington, D.C.: Brassey’s, 1997); Frank N. Schubert andTheresa L. Kraus, eds., The Whirlwind War: The United States Army in Operations Desert Shield and Desert Storm (Washington, D.C.: Center for Military History, 1995); andRichard M. Swain, “Lucky War”: Third Army in Desert Storm (Fort Leavenworth, Kans.: U.S. Army Com- mandandGeneral Staff College Press, 1997). 4. This position appears to be guiding the current review of U.S. defense policy mandated by U.S. Secretary of Defense DonaldRumsfeld.See Thomas E. Ricks, “RumsfeldOutlines Pentagon Over - haul,” Washington Post, March 23, 2001, p. 1; Tom Bowman, “Pentagon Faces Transformation,” Bal- timore Sun, March 13, 2001, p. 1A; Greg Jaffe, “Pentagon Lists Potential Cuts in About 30 Weapons Programs,” Wall Street Journal, March 26, 2001, p. A3; andMortimer B. Zuckerman, “Rethinking the Next War,” U.S. News and World Report, March 5, 2001, p. 64.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 7

izedby the air campaign? Or were they simply outmatchedby vastly superior 5 U.S. andBritish groundforces? To test the proposition that air power neutralizedthe Iraqi groundforces, I divide the argument into ªve different versions:6 (1) air power preventedthe Iraqis from maneuvering, which is critical in desert combat; (2) air power dis- ruptedIraqi command,control, communications, andintelligence (C 3I); (3) air power severedIraqi supply lines; (4) air power attritedthe Iraqi forces too heavily for them to ªght effectively; and(5) air power broke Iraqi morale. Each version makes predictions about the conduct of the Iraqi ground forces during the groundoperation. In this article I use detailedevidencefrom the ground campaign to test these ªve explanations. The evidence from the ground campaign shows that the conventional wisdom about the Gulf War is wrong. Although air power played an important role in the coalition’s victory, its role has been exaggeratedandmisunderstood. I make two primary arguments about air power during the Gulf War. First, air power was not decisive; it did not neutralize the Iraqi ground forces. At the endof the air campaign, Iraqi groundforces couldstill maneuver, andthey still hadthe C 3I, supplies, numbers, andmorale to ªght. Second, the six-week air campaign did not play a necessary, enabling role that madethe groundwar “easy” for U.S. forces. Hadthere been no air campaign, U.S. andBritish fatalities in the groundwar wouldprobably have been slightly higher. But evidence strongly suggests that with or without the air campaign, the coalition’s groundattack wouldhave ledto a rout of historic proportions. In sum, air power contributedto the coalition’s effort, but the air campaign was neither sufªcient nor necessary for the very one-sided victory. If air power did not neutralize the Iraqi army, why were the Iraqi ground forces totally ineffective during the ground campaign? The answer is that they were simply overmatchedby the U.S. andBritish groundforces. In 1991 the Iraqis ªelded a military that was mediocre even by third world standards, and during the Gulf War they were facing the most powerful military forces in the worldon terrain ideallysuitedto U.S. andBritish military strengths. Further - more, Iraq’s timing was terrible; Iraq invaded Kuwait just after the Reagan mil-

5. An obvious thirdpossibility is that a combination of air bombardmentandthe coalition’s supe - riority on the groundneutralizedthe Iraqi groundforces. I test this thirdpossibility along with the other two in this article. 6. By “air power” I mean ªxed-wing aircraft, not helicopters; helicopters are fully integrated into groundforces.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 8

itary buildupwas completedandjust before America’s post–ColdWar downsizinghadbegun. In sum, the UnitedStates andGreat Britain took the forces that they built to ªght while heavily outnumberedagainst the best divi - sions in the Soviet military, andturnedthem loose against Iraq. The Iraqi 7 groundforces never hada chance. To be clear, this article does not argue that air power was irrelevant in the Gulf War. First, air power was critical for getting combat power to the Gulf quickly after the Iraqi invasion of Kuwait. When it brieºy lookedlike Iraq might continue south from Kuwait and invade Saudi Arabia, the United States rushedcombat aircraft to the Arabian Peninsula to prevent further Iraqi at - tacks. Transport aircraft also ºew a brigade of the 82d Airborne Division to Saudi Arabia to demonstrate U.S. resolve. Because of air power, the United States was able to quickly “draw a line in the sand” and establish an early ca- 8 pability to defend Saudi Arabia. Second, during the coalition’s offensive to liberate Kuwait, U.S. aircraft successfully carriedout all of their important tra- ditional roles: They gained air superiority, ºew successful reconnaissance missions, were signiªcantly involvedin coordinatingthe actions of coalition air andgroundforces, conductedstrategicattacks on targets deepin Iraq, andre- ducedanddisruptedIraq’s military. Air power fulªlledits traditionalmis- sions; it just did not neutralize the Iraqi forces. What does the Gulf War tell us about U.S. foreign policy and procurement priorities? The conventional interpretation of the war has been usedto support the view that there has been a fundamental shift in the relative decisiveness of air and ground forces in modern combat. Modern air power, it is now widely believed, has rendered heavy ground forces obsolete. The United States—with its large, advanced air force—should have little difªculty in defeating potential opponents, as long as America commits the full weight of its air power to the ªght. This interpretation also suggests that the UnitedStates shouldcut heavy divisions in favor of lighter ground forces, tactical aircraft, long-range bomb- ers, andcruise missiles.

7. This article is part one of a larger project to determine why the Iraqi ground forces were de- featedso easily. In the secondpart, I use detailedstudiesof the major battles of the groundwar to determine what about the ground campaign was so decisive. For example, I assess the relative importance of U.S. training, groundwar technology, tactical surprise, operational surprise, andWest - ern tactics. 8. On the pace of the deployment to the Gulf and the role of air power in Desert Shield, see Winnefeld, Niblack, and Johnson, A League of Airmen, pp. 25–54; and GWAPS, Vol. 2, Pt. 1, pp. 11–21.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 9

My interpretation of the Gulf War, in contrast, paints a more complicated picture of the emerging relationship between air andgroundforces. Air power failedto neutralize the Iraqi groundforces because destroying a largely static, defensive force from the air is inherently difªcult, even in the era of information-age intelligence andprecision-strike weapons. The lesson of the Gulf War is not that air power is a weak instrument of national military power, but that the capabilities of air power against mechanizedgroundforces on the offen - sive are substantially greater than air power’s capabilities against defensive forces. The implication of my analysis for U.S. foreign policy is that air power may play a decisive role in future U.S. operations to halt an enemy’s mecha- nizedassault on a U.S. ally. It will not likely be decisive,however, if the United States or its allies needto conductan offensive to take enemy-controlledterri- tory. For example, if North Korea attacks South Korea, or if Iraq invades Ku- wait and Saudi Arabia, U.S. air assets may play a leading role in the destruction of the invading forces. But if the U.S. objective in these contingencies is to launch a counteroffensive into North Korea, or to once again evict Iraq from Kuwait, air power will be far less effective against defensively ori- entedNorth Korean, or Iraqi, forces. The force structure implications of this analysis are straightforward: If the United States envisions launching offensive operations to defeat its enemies, it will still require a balanced military that includes substantial heavy ground forces. Overemphasizing air assets may prove 9 very costly. This article is divided into six main sections. First, I lay out ªve mechanisms through which air power may have neutralizedthe Iraqi military. In the sec - ond section, I describe what happened during the four-day ground offensive. In the thirdsection, I use the history of the groundwar to test each of the ªve mechanisms. In the fourth section, I test the argument that the air campaign was a necessary “enabler” for U.S. groundforces andmadethe groundwar far

9. Maintaining an exaggeratedview of the lethality of air power may impose two types of cost on the UnitedStates andits allies. First, unduehopes for a cheap victory from the air may encourage the UnitedStates to become involvedin conºicts, only to be surprisedwhen air power is not deci - sive. At this point the UnitedStates may be forcedto choose between accepting defeator escalat - ing to include ground operations, both of which may be less attractive than staying out in the ªrst place. Second,if the UnitedStates andits allies believe that air power is more lethal than it really is, they may unbalance their military forces andeliminate (or substantially reduce)heavy ground forces in favor of air assets andlight groundforces. If a war occurs, they may pay a cost in higher casualties if they have too few heavy groundforces.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 10

less costly than it would have been otherwise. In the ªfth section, I address the counterargument that air power has grown much more lethal since the Gulf War, andthat this war therefore tells us little about the future of warfare. In this section I describe the reasons that air power was unable to neutralize the Iraqi groundforces in 1991, andI argue that the problems that coalition air power facedin the Gulf War have not yet andwill not soon go away. I con - clude this argument with a brief review of the Kosovo air campaign. Finally, in the sixth section, I describe the implications of my ªndings for the future of U.S. force structure andAmerican foreign policy.

Myths of Air Power in the Gulf War

Military analysts are nearly unanimous in their view that air power was decisive in the Gulf War. One of the best histories of this war explains the rout of the Iraqi groundforces by noting that the air campaign “hadclearly battered 10 the enemy to near senselessness.” Another excellent historical account con- cludes that “air attacks made it impossible for the Iraqis to mount an effective 11 defense.” According to air power advocate Richard Hallion, “Simply (if 12 boldly) stated, air power won the Gulf War.” Robert Pape, author of the semi- nal work on modern air power, wrote that “Iraq’s ground forces were deci- 13 mated” by the air campaign. The two most detailed studies of the air war, the Gulf War Air Power Survey (GWAPS) andthe RAND book A League of Airmen, came to the same conclusion. The authors of GWAPS concluded that “air power was the decisive factor” in the “crushing military defeat” of the Iraqis, and that “air power made [Iraqi] resistance disorganized and totally ineffec- 14 tive.” The authors of the RAND study largely agreed, arguing that “air 15 power was indeed decisive in neutralizing Iraqi forces.” Despite the agreement that air power was decisive in the Gulf War, there is no consensus on how air power accomplishedthis objective. One view is that

10. Atkinson, Crusade, p. 495. 11. Gordon and Trainor, The Generals’ War, p. 474. 12. RichardP. Hallion, Storm over Iraq: Air Power and the Gulf War (Washington, D.C.: Smithsonian Institution Press, 1992), p. 2. 13. Pape, “The Limits of Precision-GuidedAirPower,” p. 102, n. 13; andPape, “The Air Force Strikes Back,” p. 201, n. 24. 14. For the ªrst quote, see GWAPS, Vol. 2, Pt. 2, p. 104. For the secondquote, see GWAPS Summary Report, p. 117. 15. Winnefeld, Niblack, and Johnson, A League of Airmen, p. 285.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 11

air power pinneddownthe Iraqi groundforces andpreventedthem from ma - neuvering. As a result, the Iraqis were unable to defend themselves from the 16 coalition’s ºanking attack—the “left hook.” A secondpossibility is that Iraqi forces were rendered ineffective because air power disrupted their C3I. Air strikes against Iraqi command-and-control bunkers, power grids, telephone exchanges, and radio transmitters may have “blinded” the Iraqi ground forces. By the time of the groundwar, even those Iraqi units still able to communicate were afraidto doso, because any transmission invitedcoalition attention in the form of air strikes. With their communications severely disrupted, the Iraqi groundforces were unable to coordinatetheir response to the coalition attack, 17 allowing themselves to be destroyed piecemeal. A thirdversion of the air power theory is that air strikes cut Iraqi supply lines by targeting Iraqi supply dumps and, more important, destroying thousands of Iraqi trucks meant to carry supplies to the forward Iraqi divisions. By the time the groundwar began, the Iraqis were not looking for a ªght; they 18 were looking for food. A fourth explanation focuses on the attrition of the Iraqi forces in the Kuwaiti theater of operations (KTO). Statedsimply, air power may have destroyedenough Iraqi equipment andkilledenough troops

16. See GWAPS Summary Report. The authors of the report, Thomas A. Keaney andEliot A. Cohen, wrote that “the most important contribution of air power . . . anda prime reason why the ground campaign was so short andso overwhelming, was the success of air interdictionin preventing the heavy divisions from moving or ªghting effectively” (p. 116), and that air attacks “immobilized” the Republican Guard(p. 119). Winnefeld,Niblack, andJohnson, A League of Airmen, p. 153, write that “the Iraqi troops were so pinned down that it was virtually impossible to move, either day or night,” andthat “this air-supportedmaneuver dominancewas an important contributor to the co - alition success duringthe groundwar.” Thomas G. Mahnken andBarry D. Watts, “What the Gulf War Can (andCannot) Tell Us about the Future of Warfare,” International Security, Vol. 22, No. 2 (Fall 1997), pp. 158–159, also claim that the Iraqis were “pinneddown”by air power. See also Keaney, “The Linkage of Air andGroundPower in the Future of Conºict,” pp. 148–149; and Gordon and Trainor, The Generals’ War, pp. 287–288, 431.

17. Many analysts claim that air attacks on Iraqi C3I hada major effect on the outcome of the ªghting, even though they do not generally argue that these effects in themselves were decisive. See, for example, Mahnken andWatts, “What the Gulf War Can (andCannot) Tell Us about the Fu - ture of Warfare,” pp. 158–159, who write that the air campaign left the Iraqis “blinded, bloodied, and pinned down”; Winnefeld, Niblack, and Johnson, A League of Airmen, pp. 130–131, argue that

attacks on Iraqi C3I “probably signiªcantly affectedthe course of the war.” See also ibid.,pp. 152, 156–157; U.S. News andWorldReport, Triumph without Victory: The Unreported History of the Persian Gulf War (New York: RandomHouse, 1993), p. 409; andTom Clancy with Gen. FredFranks, Jr. (Ret.), Into the Storm: A Study in Command (New York: Berkley, 1998), p. 242. 18. For arguments that place substantial emphasis on the effects of air power on the Iraqi supply situation, see Winnefeld, Niblack, and Johnson, A League of Airmen, p. 156; Pape, “The Limits of Precision-Guided Air Power,” p. 113; GWAPS, Vol. 2, Pt. 1, p. 323; GWAPS Summary Report, p. 116; andKeaney, “The Linkage of Air andGroundPower in the Future of Conºict,” pp. 148–149.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 12

that when the groundwar was launched,there were not enough Iraqi forces 19 left in the theater to resist effectively. A ªfth explanation is that six weeks of air attacks broke Iraqi morale. This explanation includes elements of the other explanations: Air attacks caused se- 20 rious attrition andsupply shortages, breaking the will of the Iraqi army. The Iraqi army may still have hadthe capability to resist the coalition after the air 21 campaign, but they lackedthe will. To determine if any of these explanations is correct, we needto look at what happenedduringthe groundwar.

The Ground War

On January 17, 1991, the UnitedStates andits coalition partners launchedOp - eration Desert Storm to liberate Kuwait. For the next thirty-eight days, thousands of coalition air strikes hit targets throughout Iraq, including military commandbunkers, bridges,telephone exchanges, electrical power plants, and government buildings. In the KTO, coalition aircraft pummeled Iraqi forces andinterdictedtheirsupply lines. During the air campaign, the only ground 22 ªghting was a skirmish near the Saudi city of al-Khafji. On February 24 the

19. Ironically, despitethe signiªcant attrition that coalition aircraft inºictedon Iraqi groundforces, this explanation gets less attention than the others. For an argument that emphasizes the effects of attrition, see Pape, “The Limits of Precision-Guided Air Power,” p. 113. Keaney, “The Linkage of Air andGroundPower in the Future of Conºict,” pp. 148–149, notes the effects of attrition, though he says this was less important than preventing the Iraqi forces from maneuvering. 20. For the best evidence supporting this argument, see Hosmer, Psychological Effects of U.S. Air Operations, pp. 141–173. See also GWAPS Summary Report, pp. 107–108; Pape, “The Limits of Preci- sion-GuidedAirPower,” p. 113; Keaney, “The Linkage of Air andGroundPower in the Future of Conºict,” pp. 148–149; andLevran, Israeli Strategy after Desert Storm, p. 29. John Mueller, “The Per- fect Enemy: Assessing the Gulf War,” Security Studies, Vol. 5, No. 1 (Autumn 1995), pp. 108–111, argues that low Iraqi morale was decisive, but that Iraqi morale had been broken even before the air campaign began.

21. Note that the explanations focusing on attrition, C3I, andsupply all overlap with the morale explanation because severe attrition, lack of adequatecommandandcontrol, or lack of adequate supply couldall contribute to a collapse of Iraqi morale. I distinguishbetween these explanations in the following manner: If air power causeda collapse of Iraqi morale through any mechanism—

incessant bombardment, attrition, supply shortages, or disruption of C3I—I credit the collapse to the morale theory because air power hadshatteredthe Iraqis’ will, not their capability. I creditthe other theories only if air power destroyed Iraqi capability. For example, if air strikes damaged Iraqi

C3I to the point that Iraqi units couldnot effectively coordinatetheir actions, this wouldbe evi -

dence for the C3I explanation. But if attacks on Iraqi C3I made Iraqi soldiers feel isolated and abandoned in the theater, and made them unwilling to ªght, I code this as success for the morale explanation. 22. The al-Khafji battle was signiªcant because three Iraqi divisions attempted to advance into Saudi Arabia but were repelled by coalition air power and severely attrited. This battle demon- strates the potential lethality of U.S. air power against forces advancing in the open, but it stands in stark contrast to the inability of U.S. air power to destroy static units, and to the failure to inter-

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 13

coalition launcheda groundoffensive to liberate Kuwait anddestroy as many of the remaining Iraqi groundforces as possible. Only four dayslater, a cease- ªre was initiatedandthe war was over.

iraqi ground forces and deployments To understandthegroundwar, it is necessary to understandthegroundforces, deployments,andstrategy of the Iraqi military. On the eve of the groundwar, Iraq hadªfty-one divisionsin the KTO. Its best forces in the theater were eight Republican Guarddivisionsmannedby professional soldiersandhandpicked for their loyalty to Saddam Hussein’s Ba’th Party. They received the best train- 23 ing andequipment of all the Iraqi armedforces. The heavy divisions of the Republican Guardwere armedwith the best vehicles in the Iraqi army, princi - pally T-72 main battle tanks andBMP infantry ªghting vehicles. Nevertheless, these forces were substantially inferior to U.S. andBritish forces in both equip- 24 ment andtraining. The second-best Iraqi forces in the KTO were the nine heavy divisions of the regular army. Like the Republican Guard, these divisions were composed prin-

dict the maneuvers of the other Iraqi divisions during the ground war (see below). For a good account of the al-Khafji battle, see Gordon and Trainor, The Generals’ War, pp. 267–288; andLambeth, The Transformation of American Air Power, pp. 121–124. 23. “Heavy division” refers to any division with several hundred armored vehicles. Typically, there are two types of heavy division:armoreddivisionsandmechanizedinfantry divisions.Each generally has 100 or more tanks andother armoredvehicles. “Light divisions,”by contrast, are typically infantry divisions without substantial numbers of armored vehicles 24. The enormous technological advantages enjoyed by U.S. forces over the Republican Guard can be glimpsedby comparing the U.S. M1 tank with the Iraqi T-72. M1s were able to detectIraqi vehi - cles out to 4,000 meters andhit them—usually with their ªrst shot—beyond3,700 meters. T-72s, by contrast, had difªculty detecting U.S. vehicles beyond 1,000 meters. Furthermore, U.S. tank rounds hadno problem penetrating Iraqi tanks from any sideout to their maximum range; even the machine guns on U.S. armoredvehicles coulddestroy Iraqi T-72s from 700 meters. At that range, T-72s couldnot penetrate the frontal armor of M1s with their main guns, let alone with their machine guns. For data on the ability of U.S. weapons to penetrate the armor of Iraqi vehicles, see BattleªeldAssessment Team, “Armor/Antiarmor Operations in Southwest Asia,” Research Paper No. 92–0002 (Quantico, Va.: Marine Corps Research Center, July 1991), p. 15; U.S. Army Armor Center, “Desert ShieldandDesert Storm Emerging Observations” (Fort Knox, Ky.: U.S. Army Mili - tary History Institute, October 7, 1991), pp. 1–12. For evidence on the inability of Iraqi rounds to penetrate the front of M1 tanks, see U.S. General Accounting Ofªce (GAO), “Operation Desert Storm: Early Performance Assessment of Bradley and Abrams,” January 1992, GAO/NSIAD-92– 94, p. 24; andEzio Bonsignore, “Gulf Experience Raises Tank Survivability Issues,” Military Tech- nology, Vol. 16, No. 2 (February 1992), p. 67. For data on the range disparities between U.S. and Iraqi tanks, see U.S. GAO, “Operation Desert Storm,” p. 33; andU.S. Army Armor Center, “Desert ShieldandDesert Storm Emerging Observations,” pp. 1–10. A great overview of the breadthof the technological advantages enjoyed by the West over Soviet-designed tanks was written before the Gulf War. See Malcolm Chalmers andLutz Unterseher, “Is There a Tank Gap? Comparing NATO andWarsaw Pact Tank Fleets, International Security, Vol. 13, No. 1 (Summer 1988), pp. 23–45.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 14

cipally of professional soldiers, but their equipment and training were not as good. Heavy divisions of the regular army were typically armed with T-55 main battle tanks andolderarmoredpersonnel carriers (APCs). Although ade - quate for a thirdworldmilitary, these forces were inferior to the Republican Guardandsubstantially inferior to their U.S. andBritish counterparts. The thirdclass of Iraqi soldiersin the KTO—the frontline infantry—were conscripts deployed in thirty-four divisions near the Saudi border. They were given poor equipment andperfunctory training. Furthermore, because they were drawn predominantly from Iraq’s rebellious Kurdish and Shi’a popula- tions, who generally opposed Saddam’s Ba’th Party, their enthusiasm for 25 Saddam’s war over Kuwait was suspect. Even by third world standards, the frontline infantry divisions were extremely poor military units. Figure 1 illustrates the approximate positions of the Iraqi forces at the beginning of the coalition’s groundoffensive. Iraqi frontline infantry divisionswere deployed along the border between Kuwait and Saudi Arabia, along the Iraqi- Saudi border, and up the Kuwaiti coastline. Thick obstacle belts, including barbedwire andmineªelds,were built to buttress their positions. The frontline infantry were supportedby the heavy divisionsof the regular army, 25–50 kilometers to their rear. Behindthese were three heavy divisionsfrom the Republi- can Guard—the Tawakalna, Madinah, and Hammurabi—considered the most capable Iraqi units in the KTO. Iraq had a reasonable strategy for defending Kuwait. The frontline infantry wouldslow the coalition’s attack andchannel it into two or three breakthrough sectors. Heavy divisions of the regular army would then counterattack into the coalition’s breakthrough sectors. Finally, the heavy divisions of the Republican Guardwouldreinforce any parts of the Iraqi line being penetratedandcoun - 26 terattack the coalition forces. Although the Iraqi deployment left their western ºank exposed, the Iraqis understood how difªcult it would be for the coalition forces to navigate andmaintain supply lines across the nearly fea - tureless desert. The Iraqis apparently did not anticipate the effectiveness of the U.S. global positioning system (GPS) andother navigation equipment, which

25. “According to an MIT dissertation on the Iraqi Army, by Ahmed Hashim, the front line divisions of the Iraqi Army were 70 percent Shiite and20 percent Kurd.”Citedin U.S. News and WorldReport, Triumph without Victory, p. 404. 26. Gordon and Trainor, The Generals’ War, p. 287; andKenneth M. Pollack, “The Inºuence of Arab Culture on Arab Military Effectiveness,” Ph.D. dissertation, Massachusetts Institute of Technology, 1996, pp. 333–334.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 15

Figure 1. Iraqi Ground Force Deployments.

NOTE: The Tawakalna, Madinah, and Hammurabi Divisions are indicated by T, M, and H, respectively. Note that the Tawakalna Division was divided and deployed in two locations separated by a few kilometers; it is therefore indicated in two places. Because the focus in this article is on the Iraqi heavy divisions, Iraq’s light Republican Guard divisions are not indicated. See Central Intelligence Agency, Operation Desert Storm: A Snapshot of the Battle- field, Report IA 93-10022 (Washington, D.C.: Government Printing Office, September 1993); and Michael R. Gordon and Bernard E. Trainor, The Generals’ War: The Inside Story of the Conflict in the Gulf (Boston: Little, Brown, 1995), p. 341.

allowedU.S. forces to move off-roadacross the desertwithout getting lost. But even U.S. commanders, who had a better sense of American capabilities, won- dered whether the left hook was logistically possible. The Iraqi leadership did not expect to “beat” the coalition ground forces. Rather, they hopedthat by maneuvering their best units into the path of the co - alition’s offensive, they might be able to inºict several thousandU.S. casual -

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 16

ties. At that point, the Iraqis believed, the United States might be willing to 27 negotiate a settlement to the war.

coalition ground forces and deployments The coalition plan for the groundwar hadtwo parts. First, two U.S. Marine di - visions would attack up the middle into the teeth of the Iraqi defenses (see Fig- ure 2). The primary goals of this attack were (1) to convince the Iraqis that this was the main coalition effort; (2) to draw the Iraqi reserves—especially the Re- publican Guard—south to counterattack the marines; and (3) to advance north andhelp the coalition’s Arab partners liberate Kuwait City. The secondpart of the coalition’s plan was the U.S. Army operation. During the air campaign, the army secretly shiftedits groundforces west so that they were no longer deployed south of Iraq’s main defenses in Kuwait. From this position, the army preparedto launch its left hook. The army’s offensive was supposedto start one dayafter the marine operation. If everything went ac- cordingto plan, the Iraqis wouldrespondto the marine attack by moving their reserves south. As this happened,the army wouldrace aroundthe Iraqi ºank 28 andenvelop the Republican Guard. Two U.S. Army Corps were taskedwith executing the left hook. The VII Corps, the main effort, was orderedto ªndanddestroy the Republican Guard. A lighter force, the XVIII Corps, was deployed farther west. Its mission was to advance rapidly across the mostly empty desert, get behind the Iraqi forces in 29 the KTO, andprevent an Iraqi retreat.

day 1—february 24 The marine groundattack startedearly on the morning of February 24 andbe - gan better than U.S. planners hadexpected.The marines met little resistance as

27. Gordon and Trainor, The Generals’ War, p. 180; and GWAPS, Vol. 2, Pt. 1, pp. 111–113, 125–126. For a statement that supports this view of Iraqi strategy, see Saddam Hussein, “Speech to the Is- lamic Conference,” Baghdad Voice of the Masses, January 11, 1991, in Foreign Broadcast Information Service (FBIS)–Near East 91-009; and“Interview with Yevgenni Primakov,” Paris Europe Number One, April 28, 1991, in FBIS–Soviet Union 91-083, p. 11. Both are citedin Pollack, “The Inºuence of Arab Culture on Military Effectiveness,” p. 97. 28. Some army units, such as the 82dAirborne andthe 101st Air Assault Division, were supposed to start the same day as did the marines, but the main army effort was to begin the day after the marines attacked. 29. The VII Corps began the groundcampaign with four heavy divisions—three andone-thirdU.S heavy divisions and two-thirds of a British armored division. On the third day of the ground war, the corps was given another heavy U.S. division. The XVIII Corps, by contrast, was much lighter. Designedfor quicker movement, it hadone andone-thirdheavy divisionsplus an airborne divi - sion, a helicopter–mobile air assault division, and a medium-weight French division.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 17

Figure 2. Coalition Forces and Deployment Plans.

they advancednorthinto Kuwait. The Iraqi frontline infantry showedno inter - est in ªghting, and thousands surrendered without ªring a shot. The biggest problem for the marines on the ªrst day was ªnding a way to keep track of, 30 andcare for, the Iraqi prisoners. In fact, the marine offensive was going so well, andthe Iraqis were showing so few signs of resistance, that U.S. com - manders began to worry. If the Iraqis did not resist, the marine attack might not draw the Iraqi reserves south; instead the Iraqis might ºee north before the 31 U.S. Army hada chance to unleash the left hook. By midday, U.S. command-

30. See Charles J. Quilter II, U.S. Marines in the Persian Gulf, 1990–1991: With the I Marine Expedi- tionary Force in Desert Shield and Desert Storm (Washington, D.C.: Headquarters, United States Ma- rine Corps [USMC], 1993), pp. 75–91; Gordon and Trainor, The Generals’ War, pp. 355–363; and Scales, Certain Victory, pp. 221–222. 31. Gordon and Trainor, The Generals’ War, p. 362.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 18

Figure 3. Day 1—February 24.

ers decided to move up the army’s attack by one day, beginning its advance at dusk on February 24 (see Figure 3). The XVIII Corps made great progress on February 24. They were deployed so far west that there were few Iraqis in front of them, so the groundelement of the XVIII Corps chargedquickly aheadinto the Iraqi desert.A helicopter as - sault movedother units from the XVIII Corps 100 kilometers into Iraq toward 32 Highway 8, the main road connecting Basra with Baghdad. The VII Corps, the U.S. Army’s main force, proceeded more slowly than either the marines or the XVIII Corps. Facing nothing but open desert, the western half of the VII Corps (like the XVIII Corps to the west) chargedaheadfor the ªrst few hours of the attack. But the eastern side of the VII Corps had to breach mineªeldsandother obstacles before they couldadvanceinto Iraq.

32. Scales, Certain Victory, pp. 217–220, 254–256.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 19

Rather than allowing his forces to conduct the dangerous breaching operation at night, the VII Corps commander halted both wings of the corps until 33 dawn. At the endof the ªrst dayof groundcombat something unexpectedhap - pened. Late on the evening of February 24, only six hours after the army had launchedits surprise attack in the west, the Iraqi military began to react to the left hook. Senior Iraqi military leaders ordered two brigades of the 12th Ar- moredDivision to move west to defendagainsta possible coalition attack 34 coming from that direction (see Figure 3). The movement of the Iraqi bri- 35 gades was tracked by U.S. JSTARS surveillance aircraft.

day 2—february 25 The seconddayof the coalition attack involvedserious ªghting for the marines, but no signiªcant engagement for the U.S. Army or British forces in the left hook. The day began with a nasty surprise for the marines. Early on the morning of February 25, as the marines preparedto advancenorth, forces from two Iraqi divisions counterattacked. One Iraqi brigade (from the 5th MechanizedDivision), which hadbeen deployedinthe thick smoke arounda burning Kuwaiti oil ªeld, attacked the ºank of a marine division. Although initially caught off guard,the marines quickly recoveredandroutedthe Iraqis, 36 suffering no fatalities in the process. In the secondbattle, a brigadefrom

33. Gordon and Trainor, The Generals’ War, pp. 375–380. 34. A brigade is approximately one-third the size of a division. Some histories have argued that a brigade of the Tawakalna Division also moved on February 24. I do not believe that this is correct, but if it is true, it strengthens my argument. I beneªtedfrom discussionswith Kenneth Pollack on this point. The movement of the Iraqi reserves is described in Central Intelligence Agency (CIA), Operation Desert Storm: A Snapshot of the Battleªeld, Report IA 93–10022 (Washington, D.C.: GPO, September 1993); Scales, Certain Victory, p. 238; andSwain, “ Lucky War,” pp. 230, 233. 35. The Iraqi order to maneuver west was intercepted by U.S. intelligence. See Atkinson, Crusade, pp. 438–439. JSTARS, which stands for joint surveillance target attack radar system, is an aircraft with sophisticatedradarequipment that allows it to track the movement of vehicles on the ground over several hundredsquare kilometers. JSTARS trackedthe movement of the 12th ArmoredDivi - sion. A reproduction of a JSTARS radar picture of this Iraqi maneuver can be found in Scales, Cer- tain Victory, p. 248. 36. For details on this battle, see Press, “Lessons from Ground Combat in the Gulf,” pp. 139–143, which is basedon the two best high-resolution descriptionsof the battle: an unpublishedhistory by John H. Turner, USMC, “Counterattack: The Battle at Al Burqan,” March 1993; andCharles H. Cureton, USMC Reserve, U.S. Marines in the Persian Gulf, 1990–1991: With the 1st Marine Division in Desert Shield and Desert Storm (Washington, D.C.: Headquarters, U.S. Marine Corps, 1993), pp. 91– 96. Turner was the operations ofªcer in the regiment of the 1st Marine Division that bore the brunt of the main Iraqi attack.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 20

another Iraqi divisionattackedthe 2dMarine Division. The Iraqis were easily 37 defeated. Out west, the XVIII Corps continuedits rapidadvancethrough Iraq. U.S. light infantry units conducted a helicopter assault into positions astride High- way 8, andthe groundforce continuedto race ahead,essentially unopposed, through the desert (see Figure 4). The VII Corps advancedintoIraq andfought two minor battles. The British 1st ArmoredDivision (undercommandof the VII Corps) overran the Iraqi 52d ArmoredDivision, which hadbeen targetedduringthe air war for especially heavy air attack, anddefeateditwithout much of a fight. The U.S. 2dArmored Cavalry Regiment, in the western half of the VII Corps advance, encountered the leadbrigadeof the Iraqi 12th ArmoredDivision, the unit that hadbeen or - dered to move west the day before. Already signiªcantly attrited by air power 38 39 as it movedwest —30 of the 90 Iraqi tanks hadbeen destroyed —the Iraqi brigade was easily defeated. At the endof the seconddayof the groundwar, Iraq madetwo surprising responses to the coalition attack. First, Baghdad radio announced that the Iraqi army, having been victorious, wouldpull out of Kuwait; the Iraqi army imme- 40 diately began its retreat. Second, Iraq shifted more of its forces west to block the left hook. Two brigades of the Tawakalna Republican Guard Division, a brigadeof the 12th ArmoredDivision, the 17th ArmoredDivision, andtwo other heavy divisions of the Republican Guard—the Madinah and the 41 Hammurabi Divisions—all began moving west. As the Iraqis maneuvered, U.S. surveillance aircraft detected a column of Iraqi vehicles moving north out of Kuwait City; hundreds of coalition aircraft attackedthe Iraqi column along what wouldbe known as the Highway of Death. The air attack claimed1,400 vehicles, mostly military trucks but also some civilian vehicles. Only about 30 Iraqi armoredvehicles were among those 42 destroyed in the attack.

37. For details on the 2d Marine Division’s battle, see Quilter, With the I Marine Expeditionary Force, pp. 91–93; andDennis P. Mroczkowski, U.S. Marines in the Persian Gulf, 1990–1991: With the 2nd Ma- rine Division (Washington, D.C.: Headquarters, U.S. Marine Corps, 1993), p. 53. 38. Air power’s interdiction of this Iraqi brigade is discussed later. 39. CIA, Operation Desert Storm; Scales, Certain Victory, pp. 233–235; andClancy andFranks, Into the Storm, pp. 369–370. 40. Swain, “Lucky War,” p. 250; and GWAPS, Vol. 2, Pt. 1, p. 307. 41. CIA, Operation Desert Storm; andClancy andFranks, Into the Storm, pp. 369–370. 42. Despite the graphic pictures of the carnage, few Iraqis died along the Highway of Death. Mueller, “The Perfect Enemy,” pp. 91–92. Nevertheless, the gruesome pictures of Iraqis killedin

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 21

Figure 4. Day 2—February 25.

At the endof the seconddayof the groundwar, the battleªeldwas in great ºux. Virtually all of Iraq’s mobile forces in the Kuwaiti theater were on the move: Some were heading west to block the left hook; others were pulling out of the theater altogether. Meanwhile, the coalition forces were racing to close in on the Iraqis from the south andthe west.

day 3—february 26 The thirddaysaw the most serious groundcombat of the war. U.S. Marines fought a series of small battles outsideKuwait City andthen escortedArab co - alition forces to the outskirts of the city, so that the Arabs couldofªcially liber - ate Kuwait. For the two marine divisions, the offensive was over.

these air attacks apparently hada powerful effect on the image of the battleªeldthat crystallizedin the minds of war planners in Riyadh, leaders in Washington, and Americans in their living rooms. On the type of vehicles destroyed along the Highway of Death, see CIA, Operation Desert Storm.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 22

For the U.S. Army, the war was just beginning. Two days after the Iraqi frontline infantry surrendereden masse, anda dayafter the Iraqis announced their withdrawalfrom Kuwait, the army ªnally foundandengagedthe Iraqi Republican Guardat their new, western-orientedlines of defense(see Figure 5). On the western ºank, the XVIII Corps continuedadvancing.Its heavy groundforces crossedHighway 8, fought a series of small engagements near the highway, andseizedtwo Iraqi airªelds.The 24th Infantry Division discov - ered supply depots ªlled with all kinds of Iraqi military equipment undam- 43 agedby the air campaign. For the VII Corps, February 26 was the climactic day of the war. Throughout the day, ªve major battles were waged between elements of the VII Corps and two Iraqi divisions—the Tawakalna and the remnants of two brigades of the 44 12th ArmoredDivision. In every one of these battles, the Iraqis were on the defensive. The quality of the Iraqi defensive positions varied. Some were quite 45 good, involving reverse-slope defenses, mineªelds, dismounted infantry in trenches, andin some cases deepholes from which Iraqi tanks couldbrieºy 46 emerge, ªre, andreturn to safety. Other Iraqi positions were poor andsug- gest that some Iraqi units hadmovedinto these positions only minutes before being attackedby the coalition. The Iraqis encounteredby the VII Corps on February 26 didnotsurrender, andin many cases they fought with courage; all their efforts, however, were futile. U.S. helicopters andtanks smashedthe Iraqi armoredforces they encountered;missiles ªredfrom American helicopters, andthe main guns on U.S. tanks, routinely destroyed Iraqi armored vehicles from thousands of meters away. In the dark of night, the Republican Guard infantry made easy tar-

43. These supply dumps are described in more detail below. See Scales, Certain Victory, pp. 259, 306; andAtkinson, Crusade, p. 455. 44. Each of the ªve battles on February 26 that I call “major engagements” involvedbattalion- sizedunits or greater on both sides.An armoredor mechanizedbattalion comprises approxi - mately 50 armoredvehicles (tanks, APCs, andinfantry ªghting vehicles). Of these battles, good data on both U.S. and Iraqi forces are available for three of them. These data are summarized in Table 1. 45. A reverse-slope defensive position is one established on the rear side of a hill near its crest. One purpose of this type of deployment is to neutralize an enemy’s superior weapon range. As the enemy crests the hill, defensive forces can engage them at short range. 46. Scales, Certain Victory, pp. 261, 270, 272–273, 284–285; U.S. News andWorldReport, Triumph without Victory, pp. 338–339; Clancy andFranks, Into the Storm, pp. 398–400; U.S. Army Armor Cen- ter, “Desert ShieldandDesert Storm Emerging Observations,” pp. 2–33; andPollack, “The Inºuence of Arab Culture,” pp. 346, 348, 352.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 23

Figure 5. Day 3—February 26.

gets for U.S. andBritish tank gunners with thermal sights. Andrepeatedly, advancedarmorprotection on U.S. tanks savedAmerican lives. Iraqi rounds ricocheted off U.S. vehicles or disabled the vehicles without killing the soldiers 47 inside. Despite the Iraqis’ willingness to ªght, these battles were astonishingly one- sided. In the ªve major battles on this day, approximately 350 Iraqi armored vehicles were destroyed. Hundreds of Iraqi soldiers in the vehicles and ªghting on foot nearby were undoubtedly killed. Astonishingly, U.S. losses in these battles numberedone killedby enemy ªre andtwelve killedby friendly 48 ªre (see Table 1).

47. For a description of the technological advantages enjoyed by U.S. ground forces, see n. 24. 48. See Biddle, “Victory Misunderstood,” p. 147; Scales, Certain Victory, pp. 267–270, 272–281, 285; Atkinson, Crusade, p. 462; andU.S. News andWorldReport, Triumph without Victory, pp. 368–370.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 24

Table 1. Major Ground Engagements in Operation Desert Storm.

Force Iraq on: Ratio Battle U.S. Unit (offense/ (United Number (engaged defense States: U.S. and Date elements) Iraqi Unit meeting) Iraq) Fatalities #1* 1st Marine Division 5th ID meeting 1:1 0 February 25 (2 full-strength (2 brigades at battalions) half strength)

#2 2d Marine Division 3d AD meeting NA NA February 25 (elements) (elements)

#3 2d Marine Division 3d AD defense NA NA February 26 (elements)

#4* 2d ACR (3 troops, Tawakalna Division defense 1:1.5 1 February 26 65 armored vehicles) (96 armored vehicles)

#5 3d AD (elements of Tawakalna Division defense NA 5 February 26 1st and 2d Brigades) (number of (friendly armored vehicles fire) unknown)

#6* 1st AD (1 battalion Tawakalna Division defense 1:1 0 February 26 of 3d Brigade: 41 (45 armored armored vehicles) vehicles)

#7* 1st ID (2 battalions Tawakalna Division defense 1:1 1 February 26 of 1st Brigade; 120 and 12th AD (120 (friendly armored vehicles) armored vehicles) fire)

#8 1st ID (3d Brigade) 12th AD defense NA 6 February 26 (friendly fire) #9* 1st AD (elements of Madinah Division defense 1.5:1 1 February 27 1st and 2d Brigades; and 17th AD (110 165 armored armored vehicles) vehicles)

As the thirddayof the groundwar came to a close, andthe Tawakalna and the Iraqi 12th Armored Divisions were being decimated, the Madinah and Hammurabi Republican GuardDivisions continuedto move west into block - ing positions. Elements of the 17th ArmoredDivision deployedalongsidethe Madinah. Other Iraqi units in the KTO continued to ºee (see Figure 5).

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 25

day 4—february 27 The fourth day of the ground war was anticlimactic. The XVIII Corps continuedits advanceeast, ªghting several minor battles andpreparing to surround the remaining Iraqi forces from the north (see Figure 6). The VII Corps hadone more battle to ªght. One of its divisions—the U.S. 1st Armored Division— cresteda hill anddiscovereda brigadeof the MadinahDivision 3 kilometers in the distance.At that range, U.S. armoredvehicles couldhit the Iraqis, but Iraqi ªre couldnot reach the Americans. The U.S. tanks, supportedoverheadby at - tack helicopters, slowly advanced while shooting at every Iraqi vehicle they could see. The 1st Armored Division destroyed the brigade of the Madinah Di- 49 vision; only one U.S. soldier was killed by enemy ªre in this battle. A cease- ªre took effect early on the morning of February 28.

NOTE: The five battles with an asterisk next to the date (1, 4, 6, 7, and 9) are those for which there are good data about the forces on both sides. All force numbers are approximate and are based on author’s analysis. In the column listing U.S. fatalities, the notation “friendly fire” indicates that all U.S. fatalities in that battle were from friendly fire. ABBREVIATIONS: NA means the data are not available. ID is infantry division, AD is armored division, ACR is armored cavalry regiment, which equals approximately 1/3 of an armored division. SOURCES: The data for this table were compiled from Charles H. Cureton, USMC Reserve, U.S. Marines in the Persian Gulf, 1990–1991: With the 1st Marine Division in Desert Shield and Desert Storm (Washington, D.C.: Headquarters, U.S. Marine Corps, 1993), pp. 91–96; Charles J. Quilter II, U.S. Marines in the Persian Gulf, 1990–1991: With the I Marine Expedi- tionary Force in Desert Shield and Desert Storm (Washington, D.C.: Headquarters, United States Marine Corps [USMC], 1993), pp. 91–93; Dennis P. Mroczkowski, U.S. Marines in the Persian Gulf, 1990–1991: With the 2nd Marine Division (Washington, D.C.: U.S. Marines Headquarters, 1993), p. 53; John H. Turner, “Counterattack: The Battle at Al-Burqan,” un- published manuscript, March 1993; Robert H. Scales, Certain Victory: The U.S. Army in the Gulf War (Washington, D.C.: Brassey’s, 1997), pp. 221–222, 270, 281–285, 292–300; Rick Atkinson, Crusade: The Untold Story of the Persian Gulf War (Boston: Houghton Mifºin, 1993), pp. 466–467; Tom Donnelly and Sean Naylor, Clash of Chariots: The Great Tank Bat- tles (New York: Berkley, 1996), pp. 259, 266–272; U.S. News and World Report, Triumph without Victory: The Unreported History of the Persian Gulf War (New York: Random House, 1993), pp. 377–391; Central Intelligence Agency, Operation Desert Storm: A Snap- shot of the Battlefield, Report IA 93–10022 (Washington, D.C. Government Printing Office, September 1993); Richard M. Bohannon, “Dragon’s Roar: 1–37 Armor in the Battle of 73 Easting,” Armor, Vol. 101, No. 3 (May–June 1992), pp. 11–17; Gregory Fontenot, “Fright Night: Task Force 2/34 Armor,” Military Review, Vol. 7, No. 1 (January 1993), pp. 38–52; and W.M. Christenson and Robert A. Zirkle, “73 Easting Battle Replication—A Janus Com- bat Simulation,” P-2770 (Alexandria, Va.: Institute for Defense Analyses, September 1992), pp. A-8, 11.

49. Scales, Certain Victory, pp. 292–300.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 26

Figure 6. Day 4—February 27.

Testing the Air Power Explanations

In this section I test ªve explanations for how air power might have neutralizedthe Iraqi groundforces. These explanations are that air power (1) pre - ventedthe Iraqi forces from maneuvering, (2) severedtheir C 3I, (3) cut Iraqi supply lines, (4) attritedthe Iraqi forces, and(5) broke Iraqi morale. At the end of this section, I consider the possibility that air power neutralized the Iraqi groundforces through the cumulative effects on Iraqi maneuver, C 3I, supply, force numbers, andmorale.

did air power prevent the iraqis from maneuvering? The predictions from this explanation are straightforward: If this explanation is correct, the Iraqi forces, especially the mobile reserves, shouldhave been largely stationary duringthe groundwar. If they triedto move, this explana - tion predictsthat they shouldhave been attackedsavagely by coalition aircraft,

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 27

either destroying them or forcing them to hunker down and abandon their at- tempts to maneuver. The claim that coalition air power neutralizedthe Iraqi groundforces by pre - venting them from maneuvering is clearly wrong. During the groundwar, vir - tually all of Iraq’s mobile divisions—which comprised nearly all of Iraq’s combat power—were on the move. Of the nine heavy divisions of the regular army in the KTO, two counterattackedsouth into the marines, two moved west into blocking positions to oppose the left hook, andfour ºednorth after the general retreat was ordered. Only one (the 52d Armored Division) did not move; it was deployed very close to the Iraqi-Saudi border and was overrun within hours of the beginning of the groundwar. All three heavy divisionsof the Republican Guardmovedwest to block the path of the VII Corps. Coalition 50 air power did not pin down the heavy Iraqi divisions in the KTO. Furthermore, the Iraqis were not savagedby air power duringtheir maneuvers. More than 3,000 Iraqi armoredvehicles were on the move duringthe groundwar, andonly about 150 of these were destroyedin concentrations along the roads by coalition aircraft. More important, almost all of the vehicles destroyed along the roads were moving north to withdraw from the theater 51 rather than west to oppose the left hook. For example, on the night of Febru- ary 25, the U.S. Air Force decimated a column of retreating Iraqis along the Highway of Death; but while this was happening, all three Republican Guard heavy divisions and two heavy divisions of the Iraqi army were moving west to block the left hook. The heavy divisions moving west to ªght were hardly 52 hit as they maneuvered. Clearly, coalition air power was capable of destroying Iraqi vehicles moving along the roads; air power did this very effectively during the battle of al- Khafji andon the Highway of Death. But coalition air power failedto pin down the good units in the Iraqi military (the Republican Guard or heavy divi- 53 sions of the army) as they movedwest to ªght the VII Corps.

50. Air power did not pin down the light divisions of the Republican Guard either. The Adnan Di- vision movedwest to block the left hook; the Republican GuardSpecial Forces Division andthe BaghdadDivisionwere shiftednorth towardBaghdadtoprotect the capital, andthe al-Faw Divi - sion was sent towardBasra. See CIA, Operation Desert Storm; andPollack, “The Inºuence of Arab Culture,” pp. 344–345. 51. CIA, Operation Desert Storm. 52. Only one concentration of 30 Iraqi vehicles was destroyedalong the roadby coalition air power as it headedwest.Those were from the Iraqi 12th ArmoredDivision. They were attackedas they maneuvered west during the ªrst day of the ground war. Ibid. 53. Why did coalition aircraft fail to destroy the Republican Guard as it maneuvered west to ªght the VII Corps? This is still a mystery, andan important one, but there are at least four plausible ex -

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 28

did air power sever iraqi c3i? According to this explanation, air power neutralized the Iraqi army by cutting their C3I, effectively blinding them to the coalition’s moves and making it impossible for the Iraqis to coordinate their defense. If this explanation is correct, we wouldexpect to see an Iraqi army that either didnotreact to coalition moves, didnotreact quickly, or didnotreact well. The Iraqi groundforces shouldhave been confusedanddisorganized. There are signs that Iraqi C3I suffered from coalition attacks. When Saddam Hussein orderedthe attack on al-Khafji, he calledthe corps commanderwho wouldexecute the attack back from the KTO to Baghdadtogive him the or - ders in person. This may or may not have caused problems with the al-Khafji attack, but it is a sign that Saddam Hussein no longer trusted his C3I capability. Another sign of Iraq’s degraded C3I capability appearedduringthe ground war: The Iraqi corps commander communicated emergency orders to redeploy several Iraqi divisions without encryption, allowing U.S. intelligence to inter- cept the conversation. Despite the attacks on Iraq’s C3I system, the Iraqi senior military command- 54 ers retainedsurprisingly goodcommandandcontrol throughout the war. For example, they identiªed the left hook with remarkable speed. Only six hours after the U.S. Army launchedthis operation, the Iraqi corps commanderhada strong enough sense that something important was happening out west to maneuver two heavy brigades to new west-facing blocking positions. A day later, apparently convinced(correctly) that the main U.S. effort was coming in the left hook, he deployed his best divisions west. That he correctly identiªed the location of the main coalition attack so quickly is remarkable because Iraqi defenses were collapsing on all fronts. Given the speedwith which Iraq’s south - ern Kuwaiti defenses were crumbling, it is surprising that Iraqi leaders could tell that the marines’ attack was not the main effort. Not only did Iraq’s senior military commanders identify the left hook quickly, but they successfully ordered a reasonable response. The Iraqi reserves

planations: (1) U.S. intelligence assets did not detect the movement of the Republican Guard or did not process the data quickly enough; (2) poor weather conditions prevented coalition aircraft from attacking the Republican Guard; (3) coordination problems existed between the coalition’s air planners andthe groundcommanders,andthe fear of fratricidepreventedair attacks; and(4) old- fashioned “fog of war” prevented senior U.S. commanders in Riyadh from developing a clear picture of the chaotic battleªeld, so they never recognized that the Iraqis were establishing a coherent blocking position to oppose the left hook.

54. Hosmer, Psychological Effects of U.S. Air Operations, p. 48, notes that Iraqi C3I survivedthe air campaign in reasonably goodshape. See also Scales, Certain Victory, pp. 332–333.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 29

obeyed the order to maneuver into the oncoming coalition forces despite the risks involved. In sum, despite the weeks of air bombardment, Iraqi commanders still hadenough capability to identifythe coalition maneuver, formulate a goodresponse, communicate it to their forces, andget the desiredreaction. Iraqi commanders got their best divisions into the right place to defend against the left hook; the Iraqi military was not neutralizedby air attacks on the C 3I network.

did air power cut iraqi supply lines? There are two versions of the argument that air power neutralizedthe Iraqi groundforces by severing Iraqi supply lines. One version claims that air power degraded the Iraqi supply situation so signiªcantly that it broke Iraqi morale. The secondposits that air power deniedIraqiforces the supplies that they needed to ªght effectively. Version one of the “supply” argument is a variant of the “morale” argument, which I address later. Here I test the second argument. When the groundwar began, Iraqi supplies in the KTO were plentiful. Many Iraqi supply dumps were enormous, with thousands of individual bunkers 55 scatteredover tens of acres. Air strikes couldwhittle away at the enormous stockpiles of supplies, but even after ªve weeks of bombing, the stockpiles 56 were only slightly down. This shouldnot be surprising; the Iraqis hadªve months to amass supplies in the theater before the air campaign began. Fur- thermore, the Iraqi groundforces were in stationary defensivepositions from August 1990 through February 24, 1991, so they did not consume many of their 57 supplies. As a result, when U.S. forces advancedthrough Kuwait, andespe - cially through southeast Iraq, they overran supply dumps totaling thousands 58 of bunkers ªlledwith food,fuel, andammunition.

55. For example, one U.S. division found “an immense dump, seventy-odd square miles covered with more than a thousandstorage bunkers—enough ammunition to supply an army for many months.” Atkinson, Crusade, p. 455. At another site, the Iraqi air base at al-Tallil, there “was the center of a ten-mile-square network of deep, well-camouºaged bunkers full of weapons, ammunition, andother supplies that hadbeen reservedspeciªcally to provision andmaintain Iraq’s army in Kuwait.” U.S. News andWorldReport, Triumph without Victory, p. 376. 56. Winnefeld, Niblack, and Johnson, A League of Airmen, p. 152, write: “Supply depots were so numerous andlarge that they couldnot be eliminated;however, they were methodicallyattacked throughout the war, resulting in moderate reduction in stored materials. 57. GWAPS Summary Report, p. 99. 58. For example, on the thirddayof the groundwar, the U.S. 1st ArmoredDivision captured “more than 100 tons” of munitions at an Iraqi supply depot near the town of al-Busayya. Schubert andKraus, The Whirlwind War, p. 188. This logistics dump “contained enough fuel, ammunition, andfoodto support an entire armoredcorps.” U.S. News andWorldReport, Triumph without Vic- tory, p. 325. On the ªrst dayof the groundwar, the British 1st ArmoredDivision “overran several

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 30

Not only did the supply dumps survive the air war, but the supplies got to the Republican Guardandthe heavy divisionsof the regular army. Iraq de - ployedthe Republican Guardandseveral of the army heavy divisionsnear major supply dumps; in many cases these divisions were located on top of 59 huge complexes of supply bunkers. The proof that the Iraqi armoredand mechanizeddivisionshadadequatesupplies to maneuver andªght is that when the war started, they did maneuver and they did ªght. There is no sign that supply shortages preventedthe mobile Iraqi divisionsfrom moving where they wantedto go. Andrather than surrenderwhen they madecontact with coalition forces, the Republican Guardandthe other heavy divisions fought. The supply situation was much bleaker for the Iraqi frontline infantry. These divisions suffered from severe shortages of food, water, and other necessities. It is not possible to know whether the frontline infantry still hadthe capability to ªght when the groundwar began because they lackedthe will to resist. Whether it was air power that broke the morale of the frontline infantry—by denying them supplies, by attriting them, or by some other effect—is dis- cussedbelow in the section on morale. Air power did not neutralize the Iraqi army by severing its supply lines. Air attacks greatly reduced the ºow of critical supplies to the frontline infantry, but the heavy divisions of the Iraqi army—and all the forces in the Republican Guard—still had the food, fuel, and ammunition they needed to ªght when the groundwar began.

did air attacks attrit the iraqis too heavily for them to resist? There are two versions of the argument that air power neutralizedIraq’s ground forces through attrition: (1) attrition during the air campaign destroyed

huge logistics sites” in Iraq within a few kilometers of the Saudi border. Scales, Certain Victory, p. 245. See also Swayne, “Lucky War,” p. 244; Scales, Certain Victory, p. 259; andClancy andFranks, Into the Storm, pp. 357–358. 59. When the Tawakalna Division movedwest to block the left hook, it deployedintoan area that hadbeen preparedfor a defense.The new area hadbunkers full of supplies. When the U.S. Army’s 2dArmoredCavalry Regiment hit the Tawakalna, U.S. artillery knockedout 27 ammunition bunk - ers (Scales, Certain Victory, p. 262) containing “large stockpiles of fuel, ammunition, andother sup - plies,” including about 65 trucks (Clancy and Franks, Into the Storm, pp. 357–358). A few hours later, when the U.S. 1st Infantry Division attackedanother element of the Tawakalna andthe Iraqi 12th ArmoredDivision, it reportedthat the Iraqis were deployednextto ammunition bunkers. Secondary explosions from these bunkers proved that they were full of supplies. See U.S. News andWorldReport, Triumph without Victory, pp. 349–350; andScales, Certain Victory, p. 285. When the Madinah Division deployed west to block the left hook along with the Iraqi 17th Armored Di- vision, they both deployed next to a large Iraqi supply dump. See Scales, Certain Victory, 293; and Clancy andFranks, Into the Storm, pp. 369–370.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 31

Iraqi morale, leading them to surrender; and (2) Iraqi ground forces were attritedso heavily duringthe air campaign that they were hopelessly outnum - bered in the major battles of the ground war. The ªrst version is addressed in 60 the next section. Here I consider the second explanation. Iraq’s groundforces were heavily attritedduringthe air campaign. The best analyses of Iraqi attrition from air attacks conclude that 40 percent of Iraq’s ar- 61 moredvehicles in the KTO were neutralizedduringthe air war. This ªgure 62 probably overstates the percentage destroyed by air power, but the attrition sufferedby the Iraqi groundforces from air attacks was clearly very heavy. Despite the heavy losses, the Iraqis hadenough armoredvehicles in the KTO to defendthemselvesfrom the coalition groundattacks. There were nine major ground engagements during the Gulf War. High-resolution data are available for ªve of these, making it possible to estimate the force ratios involvedin 63 these engagements (see Table 1). One of these battles was a meeting engagement: That is, the two armies ran into each other as they movedacross the desert, so neither side was on the defensive. In the other four, the Iraqis were on the tactical defense.

60. A related argument concedes that air power did not neutralize Iraq’s ground forces through attrition, but argues that this attrition made the ground battles much more one-sided than they would have been otherwise. This argument is addressed later in this article; here I focus on the question: Didair attacks reducethe size of the Iraqi groundforces to the point that they were hopelessly outnumbered? 61. These estimates were made using the following method: Any Iraqi vehicle that neither moved nor fought during the ground war was coded as a casualty of the air war. Even if there was no sign that the vehicle was hit by an air-delivered weapon, it was credited as an air-war kill. This counting rule correctly accounts for both the direct and indirect effects of the air attacks. Vehicles directly destroyedby an air-droppedbomb wouldbe includedinthis total, as wouldany vehicle indirectly neutralizedby air power for one of the following reasons: its crew was killedin an air at - tack, the crew was scaredaway by an air attack, or the vehicle couldnot be maintainedproperly because its crew fearedan air attack. See CIA, Operation Desert Storm. 62. Although the counting rule for assessing Iraqi vehicles neutralizedduringthe air war appro - priately counts both air power’s direct andindirect effects on the Iraqi groundforces, it shouldbe notedthat this methodprobably overstates the effects of air power, because it also creditsair at - tacks for vehicles disabled during the air war by things that had nothing to do with air power. For example, Iraqi crews who abandoned their vehicles on February 24 because they feared the groundattack wouldbe scoredas an air-war kill using this metric. So a large fraction of the 40 per - cent of abandoned Iraqi vehicles may have been from non-air power effects. Forty percent then shouldbe taken as a maximum for the attrition of Iraqi groundforces duringthe air war. 63. In this section I examine only the numerical balance of Iraqi andcoalition groundforces in the major battles; I exclude the qualitative differences between these two forces. I do this to separate two different causal explanations for the one-sided rout of the Iraqis. One explanation is that the Iraqis lackedsufªcient numbers to ªght, largely because of attrition duringthe air campaign. A competing argument is that the Iraqis were hopelessly outmatchedby U.S. groundforce capabili - ties (e.g., training andtechnology). By focusing here on raw numbers—speciªcally, the quantity of armored vehicles in these battles—I distinguish between the effects of air power (which reduced the number of Iraqi vehicles) andthe effects of superior groundforces.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 32

Military professionals andanalysts have long recognizedthat in tactical en - gagements, defenders can often hold out against substantially larger attacking forces. Some analysts argue that defenders can usually stave off an enemy that is up to three times bigger; others argue that three to one is not a meaningful threshold. But most analysts agree that defenders enjoy some tactical advan- 64 tages that allow them to defend against larger attacking forces. In all four of the major battles in which the Iraqis were on the defensive, Iraq hadsufªcient numbers to ªght effectively. In one of these engagements (battle 9 in Table 1), the Iraqis were slightly outnumbered(by 1.5:1)—a ratio that should have been satisfactory for Iraq given that their forces were dug into defensive positions. But in this battle, approximately 165 American armoredve - hicles destroyed110 Iraqi tanks andinfantry ªghting vehicles, andthe Iraqis 65 killedonly one U.S. soldier. In two of these (battles 6 and7), the U.S. and Iraqi forces hadessentially equal numbers. In the other engagement (battle 4), 66 the Iraqi forces hada 1.5:1 numerical advantageover the UnitedStates. De-

64. In this article I do not rely on the 3:1 rule; I simply argue that given defensive advantages, Iraq had sufªcient numbers in all ªve well-documented ground battles to ªght effectively. On defender advantages, see John J. Mearsheimer, “Assessing the Conventional Balance: The 3:1 Rule and Its Critics,” International Security, Vol. 13, No. 4 (Spring 1989), pp. 54–89; John J. Mearsheimer, “Why the Soviets Can’t Win Quickly in Central Europe,” International Security, Vol. 7, No. 1 (Summer 1982), pp. 3–39; andBarry R. Posen, Inadvertent Escalation: Conventional War and Nuclear Risks (Ithaca, N.Y.: Cornell University Press, 1991), pp. 72–73 (especially n. 3), pp. 119–120 (especially Ta- ble 3.7). 65. The U.S. forces were from the 1st ArmoredDivision; the Iraqis includedelementsof the MadinahRepublican GuardDivision andthe 17th ArmoredDivision. The main engagement in this battle was between the 2dBrigadeof the U.S. 1st ArmoredDivision andthe 2dBrigadeof the Madinah. The U.S. forces attacked with nine tank companies in a line, totaling approximately 165 M1 tanks; they destroyedabout 110 Iraqi armoredvehicles (about 65 tanks and45 armoredperson - nel carriers). See Atkinson, Crusade, pp. 466–467; CIA, Operation Desert Storm; U.S. News and WorldReport, Triumph without Victory, pp. 377–391; Tom Donnelly andSean Naylor, Clash of Char- iots: The Great TankBattles (New York: Berkley, 1996), pp. 266–272; andScales, Certain Victory, pp. 292–300. 66. In battle 6, the U.S. 1st ArmoredDivision attackedelements of the Tawakalna Division. One battalion-sizedarmoredtask force from the 3dBrigadeof the U.S. 1st ArmoredDivision attacked elements of the 29th Brigade of the Tawakalna Division. The U.S. battalion attacked with 41 M1 tanks. Almost all the sources on the battle, including ªrsthand reports, claim that the Iraqis lost between 40 and45 armoredvehicles. See CIA, Operation Desert Storm; Donnelly andNaylor, Clash of Chariots, p. 259; andRichardM. Bohannon, “Dragon’s Roar: 1–37 Armor in the Battle of 73 Easting,” Armor, Vol. 101, No. 3 (May–June 1992), pp. 11–17. The only account that lists substantially higher numbers for the Iraqis is Scales, Certain Victory, p. 270, who reports that 150 Iraqi armored vehicles were destroyed. In battle 7, the U.S. 1st Infantry Division fought elements of two Iraqi divisions: the Tawakalna and12th ArmoredDivision. There is more uncertainty about the numbers in this battle, but the biggest engagement between these divisionsinvolvedtwo U.S. battalion-sizedtask forces from the 1st Brigadeof the 1st Infantry Division (which totaledabout 120 armoredvehicles) against 100–140 Iraqi armoredvehicles. The number of Iraqi vehicles can be estimatedfrom Scales, Certain Victory,

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 33

spite Iraq’s defensive advantages and these favorable force ratios, the Iraqis managedto kill only one U.S. soldierin these four battles, while they lost more 67 than 250 armoredvehicles. In the meeting engagement (battle 1), the Iraqis andthe marines hadroughly equal numbers of forces. Nevertheless, the Iraqis were defeated in another ex- traordinarily one-sided ªght. More than 100 Iraqi vehicles were destroyed, and no marines were killed. The Iraqis had sufªcient numbers in this battle, but 68 they were nevertheless totally ineffective. Taken together, the one-sidedness of all nine battles is shocking. Judging by force size alone, the Iraqis shouldhave won most of the major engagements. Insteadthey were defeatedinall nine, lost more than 600 armoredvehicles, 69 andkilledonly two U.S. soldiers. The Iraqis were entirely ineffective against U.S. groundforces, but air power hadnot neutralizedthem by reducingtheir numbers too far; Iraqi groundforces were simply unable to compete with the 70 better-equippedandbetter-trainedU.S. andBritish divisions.

did air power breakiraqi morale ? It is impossible to measure the morale of the Iraqi groundforces directly because morale exists in the minds of people. Morale must therefore be measured indirectly by examining behavior. More speciªcally, to assess the morale of any unit, one must evaluate the actions of the unit along two key dimensions: Did the unit follow orders,anddidtheunit ªght? Basedon these criteria, the morale of the heavy divisions of the Iraqi regular army, and especially the Repub- lican Guard, appears to have been more than adequate when the ground war began.

pp. 281–285; anda ªrsthandaccount by Gregory Fontenot, “Fright Night: Task Force 2/34 Armor,” Military Review, Vol. 73, No. 1 (January 1993), pp. 38–52. Battle 4, between the U.S. 2dArmoredCavalry Regiment andelements of the Tawakalna Division andthe 12th ArmoredDivision, is one of the most carefully studiedgroundbattles of the war. U.S. forces comprised29 tanks and37 Bradleyarmoredvehicles. The Iraqis hadapproximately 96 ar - moredvehicles, including58 tanks and38 APCs. See W.M. Christenson andRobert A. Zirkle, “73 Easting Battle Replication—A Janus Combat Simulation,” P-2770 (Alexandria, Va.: Institute for De- fense Analyses, September 1992), pp. A-8, 11. 67. Ibid. 68. See Press, “Lessons from GroundCombat in the Gulf,” pp. 139–143; Turner, “Counterattack”; andCureton, U.S. Marines in the Persian Gulf War, pp. 91–96. 69. The data on the number of Iraqi vehicles destroyed in each battle are undoubtedly imperfect. But because these engagements were so one-sided, errors in the size of the Iraqi forces are unlikely to change the conclusions. 70. For a preliminary exploration of some of the aspects of coalition groundforces that gave them such huge advantagesover Iraqi groundforces, see Press, “Lessons from GroundCombat in the Gulf.” For a conºicting view, see Biddle, “Victory Misunderstood.”

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 34

Obeying some orders requires more esprit de corps than obeying others. For example, an order to “hold position and ªght” requires greater morale from the soldiers executing it than an order to withdraw. An order to maneuver into the path of an enemy force requires an even higher level of morale. This is the order that the heavy divisions of the Republican Guard, and half the heavy divisionsof the Iraqi army, receivedandobeyedon February 25 and 26. All three heavy divisionsof the Republican Guardreceivedordersto move 71 west to block the left hook, which they obeyedimmediately. In fact, the morale of the Republican Guardwas surprisingly high, given that they hadjust endured ªve weeks of aerial bombardment that had signiªcantly reduced their forces. It wouldnot have been surprising if the Republican Guardunits had disobeyedordersandheldtheir positions or ºednorth out of Kuwait. But two of these heavy divisions (the Tawakalna and the Madinah) maneuvered and then fought against the coalition; the other (the Hammurabi) movedtoward the coalition andthen withdrew after the cease-ªre. Perhaps even more surprising, four of the heavy divisions in the regular army maneuvered to ªght the coalition ground forces. Two of them maneuvered west, as ordered, and fought alongside the Republican Guard against the left hook; the other two 72 counterattackedsouth. The four heavy divisions of the regular army that withdrew north were ordered to do so; there is no evidence that any of the heavy divisions—from the Republican Guard or the regular Iraqi army— 73 refusedto ªght. Furthermore, in the battles between the U.S. Army andthe Iraqi heavy groundforces, there are numerous reports of Iraqi soldiersªghting with nota - ble courage, especially among the Republican Guard. Dismounted Iraqi Re- publican Guardinfantry attackedU.S. tanks from bunkers andtrenches,

71. Two light Republican Guard divisions were also ordered to block the left hook, which they did. CIA, Operation Desert Storm. 72. The Iraqi 12th and17th Divisions movedwest to help block the left hook. The 3dand5th Divi - sions attackedsouth into the marines. 73. In fact, the heavy divisions of the Iraqi army that stayed and fought were, on average, attrited more heavily during the air war than those that pulled out of the theater. This conªrms that the heavy divisions that pulled out did so because of the orders we know they received, not because their morale was broken. This is based on data in ibid. and my analysis. The only heavy division of the Iraqi army that remainedstationary duringthe groundwar was the 52dArmoredDivision. It was deployed so close to the Saudi border that it was overrun almost immediately by the British 1st ArmoredDivision without a ªght. The British attack occurredafter an intense aerial bombing andªerce artillery raidagainst the 52dArmoredDivision.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 35

repeatedly hitting their targets with their ineffective handheld antitank weap- 74 ons. On several occasions, Iraqi infantry feigneddeathandthen, once by - 75 passed,jumpedup andshot at the rear of U.S. vehicles. Some Iraqi infantry jumpedonto passing U.S. tanks, only to be shot off by trailing U.S. armored 76 vehicles. Several U.S. divisions reported encountering Iraqi tanks waiting with their engines off to avoiddetectionby the thermal sights on American armoredvehicles. When the U.S. tanks passedthe Iraqi positions, the Iraqi 77 tanks wouldªre at the more vulnerable sideor rear of the American vehicles. Andin several instances, Iraqi infantry andarmor repeatedlycounterattacked U.S. armoredformations after American tanks haddestroyedentire battalions 78 of Iraqi armor in a matter of minutes. Iraqi soldiers—even the professionals—were not well trained. Their shooting was extremely inaccurate, and they were hopelessly outgunned by U.S. technology. But the evidence does not sug- gest that the Iraqis were cowards. Many of them—particularly those in the Re- publican Guard—fought and died bravely. Unlike the Republican Guardandthe heavy divisionsof the Iraqi army, the Iraqi frontline infantry meltedaway at ªrst contact with coalition ground forces. There are no reports of the frontline infantry ªghting with even moder-

74. U.S. tankers reportedthat when Iraqi antitank weapons (RPGs) hit their tanks, the weapons merely damaged the personal effects that the tank crews frequently strapped to the tops of their vehicles. Scales, Certain Victory, pp. 284–285. 75. The 2dArmoredCavalry Regiment reportedIraqis feigning deathandthen shooting antitank weapons at U.S. tanks from the rear. Atkinson, Crusade, pp. 443–445; andU.S. News andWorldRe - port, Triumph without Victory, p. 338. The 1st Infantry Division reportedIraqi infantry allowing themselves to be bypassedandthen taking shots with their RPGs at the rear of U.S. tanks. Scales, Certain Victory, pp. 284–285; andClancy andFranks, Into the Storm, p. 396. 76. See Clancy andFranks, Into the Storm, pp. 394–396; andAtkinson, Crusade, pp. 443–445. 77. The U.S. 1st Infantry Division experiencedthis with the Tawakalna. See Scales, Certain Victory, pp. 284–285. The 2dArmoredCavalry Regiment foundIraqis sitting with their vehicle engines off. U.S. News andWorldReport, Triumph without Victory, pp. 445–447. 78. The U.S. 3dArmoredDivision reportedthat infantry from the Tawakalna kept ªghting and counterattacking its positions even after their armor was destroyed. Scales, Certain Victory, p. 279. The U.S. 1st Infantry Division reportedthat Iraqi antitank infantry teams from the Tawakalna kept hitting U.S. tanks, but to no effect. Then, “throughout the night, stubborn Iraqi antitank teams re- peatedlyemergedfrom previously bypassedpositions to stalk the tanks andBradleys ....Ameri- can tankers hadlittle trouble detectingthe approaching Iraqis through thermal sights, andthey cut them down like wheat with long bursts of coaxial machine-gun ªre.” For a description of Iraqi resistance andcounterattacks, see ibid.,pp. 284–285; andU.S. News andWorldReport, Triumph without Victory, pp. 368–370. The 2dArmoredCavalry Regiment reports that there were Iraqi company-sizedcounterattacks after the Battle of 73 Easting. Atkinson, Crusade, pp. 445–447. The U.S. 3dArmoredDivision reporteda battalion-sizedIraqi armoredcounterattack early on the morning of February 27.. Clancy andFranks, Into the Storm, pp. 398–400.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 36

79 ate conviction, let alone demonstrating heroism. What is clear about the frontline infantry is that they hadvery low morale when the groundwar be - gan; what is less clear is the reason for this low morale. The best evidence in support of the argument that air attacks shattered the will of the Iraqi frontline infantry comes from Stephen Hosmer of the RAND Corporation. Hosmer reviewedtranscripts of thousandsof interviews with Iraqi prisoners of war conductedbythe UnitedStates andSaudiArabia in 1991. Most of the prisoners were frontline infantry, andalmost all of them had surrenderedwithout putting up a ªght. The majority reportedthat the air 80 bombardment had played a large role in their decision to surrender. Hosmer’s research is excellent, but other evidence suggests that air power may not have been the sole cause of the frontline infantry’s collapse. The frontline infantry divisions suffered very different levels of bombing. Some along the Kuwaiti-Saudi border were attacked by thousands of coalition sorties, including large numbers of devastating B-52 strikes, and they surrendered. Others were hit with much less intensity, and few B-52 strikes, yet they 81 also surrendered. In other words, there was considerable variation across the theater in how hardthe KurdishandShi’a conscripts were bombed.Some of the divisions were not hit hard, but the result was the same everywhere: The frontline infantry did not ªght. More generally, the intensity with which a division was hit during the air war does not correspond to that division’s willingness to ªght. The heavy divisions of the Republican Guard were hit the hardest, but they were willing to ªght. The Iraqi frontline infantry aheadof the marines, andmany of the heavy divisions of the Iraqi regular army, were hit almost as hard as the Republican Guard; the heavy divisions fought, but the frontline infantry 82 surrendered. Skeptical readers might argue that the fact that there were so few U.S. casualties in these battles between U.S. forces andheavy Iraqi divisionsproves that

79. See Quilter, With the I Marine Expeditionary Force, pp. 75–91; Gordon and Trainor, The Generals’ War, pp. 355–363; andScales, Certain Victory, pp. 221–222. 80. Hosmer, Psychological Effects of U.S. Air Operations, pp. 152–176. 81. The frontline infantry divisions far out in the west along the Iraqi-Saudi border were not hit hard during the air war because the coalition did not want to warn Iraq about the left hook, and because it wantedto maximize its effort to weaken the divisionsin front of the marines. Yet even these western Iraqi divisions also surrendered immediately. For maps showing the location of each Iraqi division in the KTO and the distribution of coalition attacks against these forces, see GWAPS, Vol. 2, Pt. 1, pp. 211, 287. 82. Ibid.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 37

83 the Iraqis did not ªght. But history shows that very one-sided outcomes such those in the Gulf War have happenedbefore, when forces from the world’s leading military power ªghts troops from the less developed world. When tactics andterrain allow the leadingmilitary power to bring its superior technol - ogy andtraining to bear, the results can be catastrophic for the locals. For example, in 1879 at the Battle of Ulundi in southern Africa, a British force using early machine guns achievedan exchange ratio in excess of 100:1 against the Zulu army, even though the Zulus reportedly fought bravely. A few years later, at the Battle of Omdurman in northern Africa, the British achieved an ex- 84 change ratio greater than 200:1. The Gulf War was not a colonial war to ac- quire empire for the UnitedStates, but from the standpointof military technology andtraining, the comparison is useful. Like the colonial troops, U.S. soldiersusedsuperior technology andtraining to decimatetheir enemy in engagements in which the enemy hadlittle chance of reciprocating. Even when 85 the locals fought bravely, they never hada chance. In sum, did the air campaign affect Iraqi morale? The answer is yes. Air power certainly contributedto the utter collapse of the frontline infantry, and it might be true that it single-handedly caused the infantry to disintegrate. But even after ªve weeks of bombing, the best Iraqi units—the Republican Guard andthe heavy units of the regular army—were willing to maneuver into the path of U.S. forces andfight. Air attacks didnotneutralize the Iraqi force by crippling their morale.

the cumulative effects of air power on the iraqi ground forces Air power did not neutralize the Iraqi army by destroying its ability to maneuver or its C3I, supply, force numbers, or morale. But perhaps air attacks weak-

83. “You can hardly march two armored divisions past each other during peacetime without pro- ducingmore casualties than those so-calledbattles,” quippeda friend. 84. Bryan Perrett, The Battle Book: Crucial Conºicts in History from 1449 to the Present (New York: Sterling, 1992), pp. 225, 292; andThomas Harbottle, Dictionary of Battles (New York: Stein andDay, 1971), pp. 208–209, 290. 85. The list of the U.S. technological advantages over Iraqi ground forces—including advanced sights that allowedU.S. tanks andAPCs to see Iraqi vehicles four times as far as the Iraqis could see U.S. forces; ªre-control computers that allowedU.S. tanks to kill Iraqi vehicles on the ªrst shot four times further than the Iraqis couldhit U.S. vehicles; thermal sights that allowedU.S. forces to see the enemy equally well at night and during the day, while the Iraqis were nearly blind in the dark; ammunition that allowed U.S. tanks to destroy Iraqi tanks from the front, side, or rear, out to 4 kilometers, comparedto Iraqi tank ammunition that couldnot penetrate the frontal armor of U.S. tanks at any range; andthe navigation equipment that allowedthe U.S. forces to advanceoff-road through the barren desert,comparedwith Iraqi forces that were anchoredto the roads—suggests

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 38

enedeach of these, andthe net effect ledto the collapse of the Iraqi forces. If this hypothesis is correct, Iraqi groundforces shouldhave been in disarray prior to their major engagements with the coalition groundforces. This explanation fails, too. Although the Iraqi frontline infantry did collapse prior to the groundwar, Iraq’s heavy divisions—thosein the Republican Guardandin the regular army—were functioning, organizedmilitary units 86 until they were engagedanddestroyedin groundcombat. As proof, when the groundwar began, Iraqi commanderssuccessfully maneuveredtheir best divisions into defensive positions in the path of the coalition’s left hook. The Iraqi soldiers in these divisions followed orders and fought against the coalition groundforces. Despite their defensivepositions andfavorable force ratios, the Iraqis hadno chance in the open desertagainst the much better trainedand well-equippedU.S. andBritish divisions.The problem that the Iraqi divisions inevitably facedwas not the effects of sustainedbombardmenton their ability to maneuver, C3I, supply, force numbers, andmorale. Their problem was that they were simply out of their league.

Did Air Power Make the One-Sided Victory Possible?

The conventional wisdom about the Gulf War is that air power neutralized the Iraqi groundforces. Above I refutedthis argument andshowedthat, at the end of the air war, the best Iraqi divisions were still ready and willing to maneuver andªght. But there is a weaker version of the air power argument that must be considered: Although air attacks did not neutralize the Iraqis, perhaps they were necessary for the one-sided coalition victory. In other words, it could be arguedthat without the air campaign, many more coalition soldiersandma - rines wouldhave been killed. The weaker version of the air power argument also appears to be untrue. Evidence from the Gulf War suggests that even if there had been no signiªcant air campaign, the groundwar wouldnot have been substantially different. There may have been a few more coalition casualties (maybe an additional 20– 200 dead), but the result would still have been a rout of historic proportions. U.S. commanders were wise to conduct the lengthy air campaign because they

that the analogy between Gulf War armoredbattles andBritish battles against the Zulus may be fully justiªed. 86. The light divisions of the Republican Guard, not the focus of this article, retained their cohe- sion too, despite heavy bombardment. CIA, Operation Desert Storm.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 39

couldnot be sure that coalition casualties wouldbe so low. But if the left hook hadbeen launchedwithout a long air campaign to soften up the Iraqi forces, the coalition groundforces wouldstill have greatly overmatchedtheir Iraqi opponents. The most signiªcant effect that the air campaign hadon the groundwar is that it substantially attritedthe Iraqi groundforces before the groundwar be - gan. At most, 40 percent of Iraqi armoredvehicles were neutralizedby the air campaign. But a close look at the groundbattles strongly suggests that these vehicles wouldnot have causedmany additionalcoalition casualties hadthere been no air war. One of the most surprising aspects of the major groundbattles in the Gulf War is that U.S. losses correlate with neither the total number of Iraqi forces in the battle nor with the force ratios between the Iraqis andthe coalition. In the ªve battles for which there is goodinformation, only two U.S. soldierswere killedby enemy ªre. In the battle in which the UnitedStates enjoyedthe biggest numerical advantage, one American died. And in the one battle in which the Iraqis hadnumerical superiority, the UnitedStates lost one soldier. The consistent pattern across all of the battles is that greater numbers of Iraqi forces simply produced greater numbers of Iraqi casualties. This suggests that hadthere been 40 percent more Iraqi tanks in these battles (i.e., hadthe air war not happened), there would have been 40 percent more Iraqi casualties during the groundcampaign. The evidencefrom the major groundbattles doesnot support the argument that many more coalition soldiers would have been killed. But suppose, despite the evidence to the contrary, that if the Iraqis had 40 percent more tanks andother armoredvehicles in each of the major ground battles, they were able to inºict 40 percent more U.S. fatalities. If we increase the number of U.S. fatalities during the ground war (the total was sixty-three) by 40 percent, the total number of U.S. soldiers and marines killed in the 87 groundwar rises to eighty-eight. Predicting exact numbers of losses in a war is impossible, but given what we now know about the one-sidedoutcomeof every groundengagement during the Gulf War, it is difªcult to support the claim that the air campaign was necessary to soften up the Iraqi forces for a successful groundoperation. The

87. Many battleªeldrelationships are nonlinear; so to be even more conservative, if 40 percent more Iraqi vehicles couldhave increasedAmerican losses by 300 percent, total U.S. fatalities in the operation from Iraqi attacks wouldhave risen to only 189.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 40

groundattack wouldlikely have gone very well for the coalition even if there hadbeen no extendedairwar.

Air Power in the Gulf and over Kosovo

The lessons from the Gulf War about military technology andthe relationship between air andgroundforces are still relevant andwill likely remain so. In this section I brieºy discuss why coalition air power was unable to neutralize the Iraqi groundforces andargue that these factors have not changed.I then support this argument with a brief description of NATO’s air campaign against the Serbian army in Kosovo. I show that the same problems that the coalition experiencedas it triedto target the Iraqi army in 1991 greatly hampered NATO’s efforts against Serbia in 1999. Coalition air power was unable to do more damage to Iraqi ground forces during the air campaign because of one key fact: The Iraqis were deployed in static defensive positions. Static forces, especially those not engaged in combat, require fewer supplies than either units on the attack or units conducting a mobile defense. They burn far less fuel, require fewer spare parts, and consume less food and water. They also require minimal communication bandwidth. A division sitting still in the desert, or even conducting a single counter- maneuver (e.g., the Iraqi response to the coalition’s left hook), requires relatively little communication among its ofªcers or between the division’s leaders and their superiors. Local commanders do not need to communicate nearly as much among themselves, or with senior commanders out of the theater, as they wouldin a more mobile defenseor an offensive operation. Thus, even if air attacks can choke off most of an enemy’s supply throughput aroundthe battleªeld, and can signiªcantly reduce its C3I bandwidth, they may have little effect on an enemy in static positions. This is precisely what happenedin the Gulf War. Coalition air power suc - 88 cessfully reducedroadtrafªc into Kuwait by 90 percent, andit hinderedIraqi C3I. But because the Iraqis were not conducting long, complicated maneuvers, andbecause the groundbattles were so short, their stockpiles were sufªcient to 89 sustain their forces. Similarly, the mission given to the Iraqi groundforces was relatively simple: Defendcurrent positions andthen maneuver west to

88. GWAPS, Vol. 2, Pt. 1, p. 281; and GWAPS Summary Report, pp. 95–96. 89. GWAPS Summary Report, p. 99.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 41

block a coalition ºanking attack. Given the simplicity of these operations, Iraq’s degraded C3I system couldstill be effective. The implication is that whenever the UnitedStates faces static, defensivelyorientedfoes, air power will have great difªculty hampering—let alone neutralizing—them by attacking supply andC 3I targets. A secondfundamentalproblem resultedfrom Iraq’s defensiveposture: Static forces are very difªcult to detect from the air, even when they are in the desert and even with information-age sensors. Static forces give few signs of their location; they emit little heat, noise, or radio trafªc. In addition, it is easy to create simple decoys that are nearly indistinguishable from stationary military forces. From an aircraft ºying at 10,000 feet, it is very difªcult to discern a tank half-buriedin the sandfrom a sheet of corrugatedmetal half-buriedin the sand. Stationary forces are both inherently difªcult to observe and even more difªcult to distinguish from simple decoys. Identifying stationary vehicles and discerning real vehicles from decoys were big problems for coalition aircraft during Desert Storm. From high alti- tude, distinguishing between “living” and “dead” Iraqi vehicles was also 90 difªcult—so much so that it led to multiple hits on the same target. Even worse, the Iraqis learnedthat U.S. aircraft couldeasily be trickedandcon- structeddecoysto mimic live targets, often putting wreckage near “living” ve- 91 hicles to make them appear “dead.” The problems that coalition air power facedin its attacks on stationary Iraqi groundforces were complex, andit appears that the UnitedStates has made little progress in solving them over the last decade. The same problems ham- 92 strung NATO efforts to destroy Serbian ground forces in Kosovo in 1999. For

90. Much has been made of the “tank plinking” in the Gulf War, in which U.S. aircraft used their infrared sensors to detect the warmth radiated by Iraqi tanks partially buried in the sand. What is not usually mentionedis that “living” Iraqi tanks create the same heat signal when warmedby the desert sun as vehicles already destroyed. Furthermore, both “living” and “dead” Iraqi vehicles were very hard to distinguish from simple metal decoys, such as sheets of corrugated metal at- tachedto short sections of thick pipe, buriedin the sand. 91. The best account of Iraqi efforts to use decoys to protect their ground forces from coalition air attacks is William F. Andrews, Airpower against an Army: Challenge and Response in CENTAF’s Duel with the Republican Guard (Maxwell Air Force Base, Ala.: Air University Press, 1998), pp. 42–46. 92. One implication of this article is that initial analyses of wars—frequently conducted before adequate information is available for careful analysis—shouldbe treatedwith substantial skepticism. This applies to the analyses of the Kosovo war on which I rely here. The problems that NATO ex- periencedduringthe Kosovo war seem similar to those that the coalition experiencedin the Gulf War, but the conclusions about the war in Kosovo are preliminary.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 42

eleven weeks NATO aircraft ºew thousands of sorties over Kosovo to disrupt the Serbs’ C3I andto attrit their forces through precision attacks, leading NATO spokespeople to conªdently assess that NATO air strikes had reduced Serbian groundforces by one-third.But after the war, it became apparent that 93 NATO hadgreatly overestimatedthe effectiveness of its attacks. In fact, NATO analysts who inspectedthe hulks of destroyedSerbian weapons report - edly found that “fewer than 20 Serbian tanks, a similar number of artillery pieces, andfewer than ten armoredpersonnel carriers” hadbeen knocked 94 out. NATO’s bombing raids against Serbian ground forces ran into many of the same problems that coalition forces facedin the Gulf War. Serbian forces were easy to hide from patrolling aircraft, and the Serbs made extensive use of 95 decoys that look like real targets from 10,000 feet. A decade has passed since the Gulf War, but the problems inherent in ªnding a static, defensively oriented army from the air, and distinguishing real targets from decoys, remain.

Implications for U.S. Foreign Policy and Force Structure

Historians, military analysts, andpolicymakers have drawnthe wrong lessons from the Gulf War, andthis may have serious implications for U.S. foreign policy andU.S. force structure. Air power is now signiªcantly more lethal against groundtargets than it was before. Almost any target that is detectedand 96 identiªed can be destroyed from the air with precision munitions. But air

93. For a detailed discussion of the various NATO estimates over time, and for alternative estimates of the damage that NATO did to the Serb army, see Barry R. Posen, “The War for Kosovo: Serbia’s Political-Military Strategy,” International Security, Vol. 24, No. 4 (Spring 2000), pp. 64–66. 94. RichardJ. Newman, “A Kosovo Numbers Game: So Where Are All the Tanks NATO Killed?” U.S. News and World Report, July 12, 1999, p. 36. See also Steven Lee Myers, “Damage to Serb Mili- tary Less Than Expected,” New YorkTimes, June 28, 1999, p. 1; “NATO’s Game of Chicken,” Newsweek, July 26, 1999, p. 59; andTim Butcher andPatrick Bishop, “NATO AdmitsAir Campaign Failed,” London Daily Telegraph, July 22, 1999, p. 1. 95. For brief descriptions of Serb decoys, see Myers, “Damage to Serb Military Less Than Ex- pected”; and “NATO’s Game of Chicken,” Newsweek, July 26, 1999, p. 59. 96. There are several important caveats to this generalization. First, badweather continues to in - hibit air attacks because most sensors still cannot “see” through clouds and because laser-designa- tors, a key component of many precision weapons, cannot penetrate clouds. GPS-guided weapons can accurately hit targets in badweather but dependonweather-sensitive reconnaissance to deter - mine the precise location of the targets. Second, some targets can be protected from air attack by locating them near civilians sites, raising the risks of collateral casualties to an unacceptable level. Third, targets can be protected from air strikes by burying them in deep bunkers. Despite these ca-

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 The Myth of Air Power in the Persian Gulf War 43

power still has limitedeffectiveness against defensivelyorientedenemy ground forces. Locating enemy ground forces in the desert is maddeningly difªcult; in forest or mountainous terrain, or in urban areas, it is even harder. And distinguishing real targets from decoys frustrated the United States in both the Gulf War andKosovo. There are many contingencies in which the UnitedStates or its allies may need to destroy a defensively oriented enemy ground force. A U.S.–South Ko- rean counterattack into North Korea; a U.S.-ledcounteroffensive into Iraq; a U.S. invasion of Kosovo, Montenegro, or Serbia; or an Israeli offensive to re- take the Golan Heights or southern Lebanon wouldprobably all require over - coming defensivelyorientedenemy groundforces. In all of these cases, air power wouldprobably have only limitedeffects on enemy defenses;success wouldhinge on the effectiveness of groundforces. At least two policy implications follow from this analysis. First, the United States andits allies shouldnot overestimate the effectiveness of air attacks and undertake offensive military operations with the expectation that air power will provide a cheap victory. When enemy ground forces must be ejected from the territory they occupy, success or failure will be determined on the ground, andthe price will dependonthe ability of U.S. andalliedgroundforces to overmatch the enemy. Second,the geography andforeign policy of the UnitedStates require that it maintain a balancedmilitary force structure. Because the UnitedStates has global military commitments, it must have a military that can deploy rapidly to defend faraway allies. Air power is ideal for this mission: It can get to dis- tant battleªelds quickly and—as al-Khafji and the Highway of Death show—it can be lethal against enemy groundforces on the move. However, because its allies are far away, the UnitedStates often joins wars late. Alliedterritory often must be recaptured, and sometimes enemy territory must be taken. For these missions, the UnitedStates needsgroundforces that can dominatethe bat - tleªeld. Unless the United States military maintains large, well-trained, and well-armedgroundforces, it will not be preparedto achieve more one-sided victories like the Gulf War. A decade has passed since the Gulf War, but the six weeks of ªghting—the long air campaign andthe short groundwar—still offer the clearest view we have of the capabilities of modern, high-technology militaries, especially when

veats, the generalization is still valid—if an enemy target is identiªed and is within range of precision-strike munitions, it can usually be destroyed.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021 International Security 26:2 44

they are pittedagainst thirdworldopponents. Air power is a more powerful tool of national policy than it once was, but it still does not dominate the bat- tleªeld. Understanding the relationship between air and ground forces is necessary to make soundforeign policy decisions,to create a force structure that can meet the challenges that the United States will face in the coming decades, andin the end,to save American lives.

Downloaded from http://www.mitpressjournals.org/doi/pdf/10.1162/016228801753191123 by guest on 24 September 2021